DESCRIPTION (Adapted from applicant's description): Applicant: George A. Macones, M.D., M.S.C.E. is the Principal Investigator and Candidate for this award. Dr. Macones is an established clinical researcher, who has been successful in obtaining federal grants for clinical research. He has a track record of mentoring young investigators, including Master's trainees in Clinical Epidemiology, Maternal-Fetal Medicine Fellows, OB/GYN residents, and medical students. Dr. Macones' goals are to make important contributions to the clinical care of pregnant women, as well as to be an integral part of the training of the next generation of patient oriented researchers in obstetrics and gynecology. Research Plan: To use a randomized clinical trial to test the hypothesis that indomethacin, when used for adjuvant tocolysis in preterm labor (i.e., in those who are failing to respond to conventional first-line therapy), can decrease the rate of major neonatal complications and death. The proposed research project will use a double-blind, parallel, randomized controlled clinical trial design to compare the efficacy of indomethacin to placebo for adjuvant tocolysis in pregnancies less than 30 weeks gestation. Patient will be recruited for this study from the Labor and Delivery wards of the Hospital of the University of Pennsylvania, the Pennsylvania Hospital, the Thomas Jefferson University Hospital, and Christiana Care Health Services. Patients who "fail" first-line therapy with magnesium sulfate will be evaluated for possible enrollment into this randomized clinical trial. Those who give informed consent will be enrolled into the study. This study calls for the enrollment of 205 patients in both arms. Patients will be randomized to receive either indomethacin rectal suppositories (50mg every 6 hours) or identical appearing placebo suppositories for 48 hours. After completion of treatment with either placebo or indomethacin, patient care will be largely at the discretion of the attending physicians. Many prior studies of tocolysis have used variable periods of delay in delivery as the primary outcome measure. We believe that this is not the appropriate primary endpoint, since the ultimate goal of tocolysis is to improve neonatal outcome. Thus, although delivery delay will be analyzed as a secondary endpoint, our primary outcome (on which our sample size is based) will be based on major neonatal morbidity. Mentoring Plan: Dr. Macones has established himself as a mentor to young investigators interested in clinical research, and has spent a large portion of his time on this activity. He will continue this activity over the course of this proposal, and will focus on students in the Master of Science program in Clinical Epidemiology at the University of Pennsylvania, as well as on fellows who receive support form the NIH sponsored training grant in Reproductive Clinical Epidemiology or in the new track in Patient Oriented Research.